Media Crisis: AI Bots Steal Content and Undermine Revenues

The media crisis is being accelerated by artificial intelligence (AI) bots, which not only steal editorial content to provide quick and synthesized responses to users of chatbots like ChatGPT, Perplexity, or Gemini, but also, by doing so in real-time, have a more immediate and direct impact on the revenues of information companies. In the second half of last year, AI bot activity increased by 300%, and media is the second-largest sector where they are concentrated, according to a study by Akamai. A large part of this traffic not only aims to train AI models but also to steal media content in real-time to provide breaking news summaries through their chatbots. This has caused a 96% drop in referral traffic. Although chatbots may reference some of their sources, they do so without the consent of the publications, exposing the intellectual property theft of these models. Even when sources are cited, the problem of reduced traffic is not solved. 'AI bots are undermining revenue mechanisms such as paywalls, subscription models, and advertising-based income, on which the publishing industry depends,' according to the Akamai report. There are two types of bots that dominate online traffic: crawlers, which capture information to train AI models, and scrapers or collectors, which access real-time information from news portals to give it to chatbot users, to the detriment of the intellectual property rights of media companies. In a context where the internet is increasingly dominated by machines—according to Cloudflare's radar, a third of global traffic is generated by bots—scrapers have multiplied and now account for a quarter of all AI traffic on media platforms. According to a report by the platform Human, automated internet traffic is growing eight times faster than human traffic; last year, 95% of AI-generated activity was concentrated in three sectors: retail and e-commerce, real-time content distribution (streaming), media, and tourism. 'These are the vertical sectors where structured and frequently updated data have the highest commercial value for AI products, and where they promise the greatest utility for end-users,' explained Human in its report on the state of AI traffic and cyber threats for 2026. Ultimately, this leads to a highly concentrated market. OpenAI's bots account for about 69% of the AI traffic observed by Human; Meta-ExternalAgent contributes an additional 16%, and Claude from Anthropic accounts for about 11%. In the publishing sector, between July and December 2025, OpenAI, Meta, and ByteDance were the three main providers or owners of AI agents that affected media companies, according to Akamai. 'AI bots are undermining primary revenue streams, such as advertising and subscriptions, while increasing infrastructure costs and decreasing brand visibility,' pointed out Patrick Sullivan, Akamai's director of technology and security strategy, in a press release. Beyond cases that have led to reputational damage from the use of AI in media, the resistance of media companies seems to be an overwhelmed attempt against the power of large tech firms. However, users continue to place greater credibility on journalism assumed to be done by humans. Distrust: A Pew Research Center survey conducted in the summer of 2024 indicates that half of U.S. adults believe AI will have a negative impact on the news they receive (50%), 10% see it as positive, 23% anticipate a mixed effect, and 16% are unsure. On the other hand, 41% of U.S. adults state that AI would write worse news than journalists, 19% think the opposite, and 20% consider its performance would be similar. Likewise, six out of ten people expect AI to decimate media jobs in the next two decades. What 92% of those surveyed by the Pew Research Center agree on is the concern about the possibility of receiving inaccurate information from AI.